This proposal is to continue studies to elucidate the detailed molecular mechanisms by which certain antitumor antibiotics damage DNA and interfere with DNA function. Not only will this work contribute to our understanding of how anticancer agents work, but, hopefully, should lead to the development of drugs with increased potency and selectivity for neoplastic cells and lowered toxicity for the host. Important byproducts will be the uncovering of novel mechanisms of DNA damage and repair and of mutagenesis, and the development of new probes and tools for the study of gene structure and regulation. It is also expected that new insights on how radiation sensitizers work will be forthcoming. Members of a family of radiomimetic protein antibiotics, in particular neocarzinostatin (NCS), will be te focus of this investigation. Earlier work has shown that NCS contains a labile nonprotein chromophore of unique structure that possesses all the biological activity, binds to DNA by intercalation, and damages DNA when activated by thiol by selectively attacking C- 5' of deoxyribose of mainly thymine residues to produce, in the presence of 02, strand breaks with nucleoside 5'-aldehyde at the 5'-end and phosphoryl at the 32-end, and base release with the formation of apurintic/apyrimidintic sites, and in the absence of 02 covalent NCS chromophore-deoxyribose adducts. A similar mechanism occurs in NCS-treated mammalian cells. Studies will be performed 1) to determine the role of microheterogeneity of DNA structure in base sequence-dependent binding of NCS, using chemically synthesized oligodeoxynucleotides of defined sequence, 2) to determine structure-function relationships in the NCS chromophore by chemical modifications, 3) to obtain more detailed information on the molecular mechanism of DNA damage by chemical characterization of damage products (deoxyribose fragments, NCS chromophore-deoxyribose adducts, spent NCS chromophore, etc.), 4) to clarify how nitroaromatic radiation sensitizers, such as misonidazole, substitute for 02 in producing DNA damage, and 5) to characterize possible new enzymatic mechanisms of repair of NCS-inducd DNA damage. These studies will be extended to include other related protein antibiotics, notably auromomycin, that have different attack site specificities and produce different DNA damage products.